Hospitalizations and Mortality Following Skilled Nursing Facility Admission by Opioid Use Disorder Status.

Abstract

Prior studies have documented rising rates of referrals to skilled nursing facilities (SNFs) for hospitalized patients with opioid use disorder (OUD). However, the care transitions and survival of people with OUD who enter SNFs have not been evaluated. We examined differences in hospital readmissions and mortality between individuals with and without OUD discharged to SNFs. Using 2016-2020 100% Medicare inpatient claims, we identified acute hospitalizations with discharge to SNFs. We matched each beneficiary with OUD with up to 5 without OUD based on age, sex, low-income subsidy status, and residential county. Outcomes were hospital readmissions and all-cause mortality within 180 days following hospital discharge. Inverse-probability weighting (IPW) covariates included demographics, state of residence, year of SNF admission, intensive care use, Gagne comorbidity score, and conditions associated with SNF admission or that disproportionately affect people with OUD. Unadjusted and IPW-adjusted risk differences were calculated. There were 30,922 fee-for-service beneficiaries with OUD and 137,454 matched beneficiaries without OUD, with a mean age of ~71 years. Of those with OUD, 5.3% had evidence of receiving medications for OUD. In unadjusted analyses, beneficiaries with OUD had higher readmission risk (44.5% vs 27.9%) and comparable mortality risk (17.8% vs 16.5%) relative to beneficiaries without OUD. After adjustment through IPW, there were minimal differences in mortality; however, beneficiaries with OUD remained at a greater risk for readmissions than those without OUD. Hospital readmissions were dramatically higher among Medicare beneficiaries with OUD than those without OUD, suggesting that important gaps in OUD care exist in SNFs.